Filter
Exclude
Time range
-
Near
what is the difference between ralphloop or /goal?just cronjob?
30
Ran my first #ralphloop yesterday, fixed a bug, very cool.
4
40
codex君 もう38時間作り続けてる 「PLAN.mdに沿ってMVPではなく本番レベルまで完璧に実装して」 ってgoalコマンドで設定しただけやのに RalphLoopとかもいらんかったんやね
うーんgoalコマンド使ってないとかかな? 昨日投げた重めのサービスを1から本番水準まで作るようgoal設定したらもう12時間以上作り続けてる
2
4
1,622
ちなみに俺はRalph-loopの進化版のMayweather-loopのほうがすき。 てかRalph loopがそもそも前からあったからあんまそこまで驚くほどでもって感じ。使用用途の使い分けやけど個人的にはRalphloopがチートかな
実際/goal系のプロンプト磨きに時間使っててもしゃーないんよなーwwwもっとレバレッジの効く仕組みに時間とトークンを投下しないとそれこそ質とスピードあるやつには勝てない
2
165
These are all great but this one I actually LOL'd can't wait to slack the whole team that I automated gary with a ralphloop
May 11
The Unethical Guide to Surviving AI Layoffs
1
3
307
RalphLoopのようなアプローチだと評価は同一セッション内ではなく独立で行う、が鉄板なのでgoalとは相性が悪そう。goal使うならcodex exec呼んで評価させろ、みたいな工夫が必要。 ポン出し神ゲーみたいな動画がTwitterでたくさん流れてくるけどそんな簡単じゃないよ、ということで。
Codex Imagegen, 手動でスプライトアニメはいい感じに作れるんだけど、/goalで自動化量産した場合はどうなるか実験。Godotを使って1はスプライト、2はスケルトン。 スクリーンショット評価は入れてるんだけどgoalは単一セッションでお手盛り評価になってしまうので、定性的なタスクは苦手ですねー
3
790
May 1
Replying to @dywfjyie2489
现在的codex能力很强,有两种方案 1. 你做一个很长的spec,我昨天做了一个跑了8个小时 2. 使用类似ralphloop的skill
2
107
This guy is absolutely mad! We spent nearly three hours discussing just how good ouroboros is; it looks set to become a powerful agent OS. And seeing it running so smoothly on the Hermes Agent really got my heart racing. I’ve had enough sleep for tonight. It’s an evolved version of Ralphloop. It’ll be uploaded to the podcast soon, and I’ll make sure to include English subtitles as well!
1
1
3
289
今日時点でのAI中心の仕事環境 1. インターフェース:MacBookPro Cmux 2. 脳:Codex(cc) 3. 情報ソース拡張:Clawl4ai、xurl 4. 持ち運び:Tailscale Moshi 5. ループ:MacStudio Ralphloop 監視自律改善 6. 管理:GitHub
自宅でホスティングしてる自分だけが使うサービスをtailscale使って外出先でも触れるようにしてて便利な暮らしを実現してるんだけど、tailscale 無料 なぜ という気持ちが湧き上がってくるな。
1
6
662
Replying to @koltregaskes
They can work if you put them in a loop (ralphloop being an example), but don’t count on the output being good or legible
1
2
378
folks; heading to SF again shortly. i’m opening up @heavybit devguild. join me and a sharp lineup of infra founders and engineers in SF at luma.com/write-only-code-sum… use the coupon code “ralphloop” for 20% off.
5
1
33
2,987
Built a spec-driven AI Skill: “Ship-it-Ralph”. Give it an idea in one sentence → working app out. Runs the Ralph loop: spec → tasks → tests → code. Use in GH Copilot, Cursor, Claude Code, or Antigravity. github.com/swamichandra/ship… #specdrivendevelopment #ralphloop #skill
2
173
ArboNOVA: Patent–Molecule matching loop We’ve been experimenting with an agent that maps molecules → prior art using only open data tools Benchmark: ~1500 molecules across ADHD-related patents (since 2012) In ~12 hours: 18 iterations of the loop → Best hit rate: 85.4% How this is usually done: Pharma intelligence teams expensive proprietary databases manual workflows even conference attendance Early, but promising. Moving one step closer toward automating drug discovery and identifying which molecules are most strategic to advance in the wet lab. Based on @const_reborn (github.com/unconst/Arbos) and @karpathy autoresearch framework #Bittensor #SN68 #ralphloop #agents #DrugDiscovery #Desci #DeAI
9
36
127
27,249
Mar 4
Replying to @chiefofautism
/ralphloop
4
252
SWE-searchをralphloopのようなセッションループに組み込んで探索検討させるのを、ゲームの面白さの模索に活用するのはどうかなって考え中。 セッションループは触っててめちゃくちゃ面白い。
Feb 22
Claude Codeでralph-loopのようなロングランを実行する時には、Specの作り込みは重要。コンテキストに残すのではなく外部永続ドキュメントとして用意するアプローチは個人的も有効だと思う。 How I Use Claude Code | @boristane boristane.com/blog/how-i-use…
4
1,113
原本以为今年的假期只会用来捣鼓ai. 没想到用clawdbot, seedance, RalphLoop搭了工作流后还有时间 算是很意外地在春节期间读完了两本书,迈克尔·刘易斯的《大空头》和孙宇晨的《这世界既残酷也温柔》 我其实是一个阅读非常大量的人,而且读的很快,专注度也在线,有时还能不间断地读上一两个小时
3
294
# The Rise of the Autonomous Loop: Why Moltbot and Ralph Wiggum Are Dominating X 🦞🔄 In the last 48 hours, "AI Twitter" has shifted dramatically. We've left simple chatbots behind and entered the era of the **Autonomous Agent Loop**. Two names lead this shift: **Moltbot** (formerly Clawdbot) and the **Ralph Wiggum loop**. Here's why they're dominating timelines and demand attention. --- ### 1. The End of the "Prompt-and-Wait" Era For two years, AI use meant typing a prompt and waiting. Moltbot and Ralph patterns eliminate that. They run on a **continuous execution loop**. You assign a **mission**. The agent then: - Executes a command - Observes errors - Edits code - Retries automatically This looping explains devs posting logs of agents running for 20 minutes while they relax. Productivity compounds exponentially. ### 2. Moltbot: The Lobster That Shed Its Skin 🦞 Formerly **Clawdbot**, Moltbot rebranded recently (after a nudge from Anthropic). The "molt" became a meme for the software's evolution. Moltbot dominates with its **local-first, full-access runtime**. It runs in your terminal with file system access, browser control, and deployment pipelines—an "OS for agents." ### 3. The Ralph Wiggum Loop: Pure Chaos Meets Pure Code 🔄 Named after the Simpsons character in his own world, the **Ralph Loop** is an autonomous recursive coding pattern. Pioneered by @GeoffreyHuntley, it enables "learning by doing" in real time. It's the "I'm in danger" meme as high-speed development: loud, visible, and producing code faster than humans. ### 4. Why X Is the Battlefield X hosts these agents as **native participants**: - Posting progress updates - Replying to repo bug reports - Sharing terminal logs as proof Posts from accounts like @openclaw come directly from agent loops completing tasks—no marketing team. --- ### The Verdict We're seeing software development's "Industrial Revolution." Moltbot and Ralph loops aren't tools; they're **digital workers**. If you're not running a loop, you're competing against machines that never sleep, never bore, and retry endlessly until code works. **Welcome to the Molt. 🦞🔄** #AI #Moltbot #RalphLoop #AutonomousAgents #Coding #TechNews
3
529